Statement of the Problem: Bullying in the early elementary years is highly prevalent, and there is growing recognition that bullying is a critical antecedent of more serious mental health disorders including youth violence. Few interventions focus specifically on bullying, and those that do are typically not comprehensive, failing to address the peer relationships of the bullies or victims, relationships which are being found to have considerable influence on long-term mental health outcomes. Purpose: This project would lay the groundwork for gathering experimental evidence on the effectiveness of an adaptation of a "best practices" prevention program (Early Risers) that features a strategic peer affiliation and peer coping skills component designed to reduce aggression, bully problems, and victimization. The primary goal is to assess the feasibility, acceptability, and safety of the new program component and to gather data to estimate parameters needed in planning a full scale efficacy trial. A secondary goal is to evaluate the potential effects of the program in normalizing the developmental trajectories of the at-risk participants. An important aim of the current proposal is to systematically study the effects of the program on low-risk peer mentors. Methods: Aggressive or withdrawn kindergarten and 1st graders (n = 60) will be paired with 60 nonaggressive peer mentors as buddies. They will receive 15 months of intervention, including two summer programs, an after-school program, and (for the parents of aggressive children) a family skills and support intervention. An additional 40 aggressive or withdrawn children and 40 well-adjusted children will serve as no-intervention controls. A comprehensive assessment battery including child, parent, and teacher informants, direct observation, and peer sociometric instruments will be completed at baseline, twice during the intervention, and twice in a follow-up year. Four dimensions of fidelity (exposure, adherence, quality of delivery, and innovation) will be measured for each of five intervention components. General linear mixed modeling will be employed to analyze the data. Implications for Prevention and Relevance for Public Health: This project will provide a greater understanding of the potential of a comprehensive approach to the prevention of youth violence and victimization, which are implicated in numerous later mental health problems.